as the Sulphate and as the Oxide. 



211 



ture was boiled, and the solution of the purified and neutral 

 manganous chloride was filtered from the excess of undissolved 

 carbonate. Definite portions of this solution were precipitated 

 with silver nitrate, and from the weight of the silver chloride 

 thus obtained the amount of manganous chloride present was 

 calculated. Portions of the solution thus standardized were 

 drawn, for our experiments, from a burette into a weighed 

 platinum crucible, sulphuric was added in amount more than 

 equivalent to the manganese, the solution was evaporated on 

 the water-bath until the water was removed, and then, sup- 

 ported by means of a porcelain ring, or triangle, within a 

 larger porcelain crucible used as a radiator so that the bottom 

 and walls of the one were distant from the bottom and walls 

 of the other by an interval of about l cm , the crucible was 

 heated more strongly. The outer porcelain crucible may be 

 heated over a good Bunsen flame to a red heat without risk of 

 overheating the manganese sulphate within the inner crucible, 

 and the ignition may proceed as rapidly as is consistent with 

 the avoidance of mechanical loss by spattering. The results 

 obtained by treatment of equal portions (50 cm3 ) of the same 

 solution are given, together with the results of standardizing 

 the solution by precipitation with silver nitrate, in columns A 

 of the following table. In the other columns are given com- 

 parative results got in the treatment of equal portions of sev- 

 eral other solutions employed subsequently in other work. 



MnS0 4 



calc'd from 

 AgCl found 

 in 50 ira3 of 

 solution A. 



grm. 



MnS0 4 



found by 



treatment 



of 50 cm3 of 



solution A 



with H 2 S0 4 . 



grm. 



A A 



(1) -3518 (1)0-3513 



(2) -3512 (2)0-3514 



(3) 0-3518 



MnS0 4 found by treatment of 50 cm3 of 

 various solutions with H 2 S0 4 . 



B C D E 1 F I G- 



(1) 0-3100 (1) 03256 (1) 0'3534 (1) 0-3524(1) 0.3355 (1) 0-5475 



(2) 0-3104 (2) 0\3254 (2) 0'3543 (2) 0"3520 (2) 0*3357 (2) 0-5476 



(3) 0-3096 



These results show plainly that the process of estimating 

 manganese in the form of the anhydrous sulphate is both sim- 

 ple and accurate. 



The estimation of manganese as the manganosomanganic 

 oxide Mn 3 4 , has been so frequently criticized unfavorably that 

 the method may be said to have passed from very general use 

 excepting in certain cases in which the directness of the process 

 is a temptation to incur the risk of some uncertainty. The 

 production of the other oxides of manganese in definite condi- 

 tion is thought to be even more uncertain. Manganese dioxide, 



